Perf: 提升可视化图片清晰度(max_h 300/400->600, JPEG quality 85->95)

Fix: 修复历史图片计数不更新和历史面板显示暂无历史用户的问题

Fix: 修复首张文本场景图重复OCR问题，更新关于页技术栈徽章

Revert: 回滚ONNX(HF Spaces上反而更慢), 移除可视化文字叠加

Perf: PaddleOCR ONNX Runtime 加速，文本场景提速 ~20%

Perf: PP-OCRv3 + fast scene classify for CPU speedup

Fix: UnboundLocalError classify_scene caused by local import shadowing global

Fix: background stays fixed, remove parallax scroll on about page

Fix: clean Qdrant cache when history empty, remove retest button

Fix: add /cover.png endpoint for static file serving

Fix: add /cover.png endpoint for static file serving

Fix: demo skip classify_scene, fix unpack error, fix cover path, update track name

Add cover bg, scene mode selector, single image upload, demo hides upload, about page format

Glassmorphism UI, about page, real demo images, skeleton loading, parallax

Add demo scene button, thread config, user expansion

Fix hang: cache images in memory before delete, use 2 workers, vis from cache

Remove text area below text scene vis, keep only side-by-side images with labels

Remove Chinese keyword text from text vis image, keep in frontend HTML only

Fix: remove rect boxes from vis, pass per-image crop_offset, only draw lines

Fix: patch coord mapping with CenterCrop offset, PIL for Chinese text in text vis

Fix visualization: build before delete, add text scene vis, click-to-show

Fix: pass patch_match_info through evaluator to enable visualization

Add: DINOv2 threshold 0.5, patch match visualization, history image management

Fix paddlepaddle version to 2.6.2